The acoustics and acoustic behavior of the California spiny

lobster (Panulirus interruptus)

S. N. Patek,a L. E. Shipp, and E. R. Staaterman
Department of Integrative Biology, University of California, Berkeley, California 94720-3140
Received 15 October 2008; revised 9 February 2009; accepted 18 February 2009
Numerous animals produce sounds during interactions with potential predators, yet little is known
about the acoustics of these sounds, especially in marine environments. California spiny lobsters
Panulirus interruptus produce pulsatile rasps when interacting with potential predators. They
generate sound using frictional structures located at the base of each antenna. This study probes
three issuesthe effect of body size on signal features, behavioral modification of sound features,
and the influence of the ambient environment on the signal. Body size and file length were positively
correlated, and larger animals produced lower pulse rate rasps. Ambient noise levels 149.3 dB re
1 Pa acoustically obscured many rasps 150.4 2.0 dB re 1 Pa at distances from 0.9 1.4 m.
Significantly higher numbers of pulses, pulse rate, and rasp duration were produced in rasps
generated with two antennae compared to rasps produced with only one antenna. Strong periodic
resonances were measured in tank-recorded rasps, whereas field-recorded rasps had little frequency
structure. Spiny lobster rasps exhibit flexibility in acoustic signal features, but their propagation is
constrained, perhaps beneficially, by the noisy marine environment. Examining the connections
between behavior, environment, and acoustics is critical for understanding this fundamental type of
animal communication. 2009 Acoustical Society of America. DOI: 10.1121/1.3097760
PACS numbers: 43.80.Ka WWA Pages: 34343443

I. INTRODUCTION reviewed in Chivers et al., 1996. In either case, the propa-

Surprisingly few studies have examined the acoustics of
antipredator sounds in the marine environment even though
these sounds have been observed across a wide range of taxa
in an array of marine habitats. For example, when interacting
with intruders or potential predators, nephropid lobsters vi-
brate antennal muscles Mendelson, 1969; Henninger and
Watson, 2005, ocypodid and pagurid crabs stridulate
Guinot-Dumortier and Dumortier, 1960; Field et al., 1987,
astacid crayfish squeak with their abdomen Sandeman and
Wilkens, 1982, mantis shrimp rumble using muscles at-
tached to their carapace Patek and Caldwell, 2006, and fish
produce a myriad of sounds using mechanisms from muscle
contractions to stridulatory jaws Fish and Mowbray, 1970.
Antipredator sounds have been shown experimentally to de-
ter predators Alexander, 1958; Masters, 1979; Lewis and
Cane, 1990; Sargent, 1990 and function through a variety of
mechanisms, including the effects of startle, warning or
predator memory enhancement Gittleman and Harvey,
1980; Gamberale and Tullberg, 1996; Speed, 2000; Sherratt
and Beatty, 2003; Ruxton et al., 2004; Caro, 2005.
There are two key factors when examining antipredator
sounds: signal propagation through the environment and the
effect of, or information contained in, the signal features. In
terms of signal propagation, the emission of sound has the
potential to attract other predators to the scene; this may
make the situation for the prey item even more dangerous or,
alternatively, it may increase conflict between the predators
and thereby increase the odds that the prey item escapes
a
Author to whom correspondence should be addressed. Electronic mail:
patek@berkeley.edu
gation of antipredator signals through the environment
given the ambient background noise and the effects of the
physical structure of the habitatis an important factor in
how the signal functions.
Antipredator signal features are relevant to their function
and performance in deterring predators. Most antipredator
signals, whether acoustic, chemical, or visual, capitalize on
being noxious or generally startling to either deter the preda-
tor or trigger a predators memory that the prey item is not
palatable Edmunds, 1974; Ruxton et al., 2004. Further-
more, it is generally advantageous for the prey to appear as
threatening as possible by exaggerating size e.g., eye spots.
Although these general principles of antipredator signal de-
sign are widely accepted, the features of acoustic antipreda-
tor signals are rarely characterized.
Most spiny lobster species Palinuridae, including the
California spiny lobster Panulirus interruptus, generate an-
tipredator sounds called rasps Parker, 1878; 1883; Lind-
berg, 1955; Moulton, 1957; George and Main, 1967; Meyer-
Rochow and Penrose, 1974; Smale, 1974; Meyer-Rochow
and Penrose, 1976; Mulligan and Fischer, 1977; Patek, 2001;
Patek, 2002; Patek and Oakley, 2003; Latha et al., 2005;
Patek et al., 2006. Documented for over a thousand years
Athenaeus, 300, these sounds are produced when spiny
lobsters are handled by potential predators. The rasps func-
tion to deter predators; spiny lobsters that have been silenced
i.e., the sound-generating apparatus has been disabled are
attacked more frequently and with greater success than spiny
lobsters with intact sound-producing structures Bouwma
and Herrnkind, 2004; Bouwma, 2006; Bouwma and Hernn-
kind, 2007.

3434 J. Acoust. Soc. Am. 125 5, May 2009 0001-4966/2009/1255/3434/10/$25.00 2009 Acoustical Society of America

FIG. 1. The sound-producing anatomy of a California spiny lobster Panu-
lirus interruptus. A plectrum is found at the base of each antenna and rubs
over a file beneath each eye. Sound is produced when the plectrum slides
posteriorly arrow over the file. Adapted from Summers, 2001.
The stick-slip frictional mechanism of spiny lobster
sound production is unusual in the biological world, and the
paired structures potentially yield flexibility in signal fea-
tures. Analogous to bowed-stringed instruments, spiny lob-
sters produce pulses of sound through stick-slip frictional
interactions between the plectrum and file surfaces such that
the plectrum sticks and slips due to friction as it is pulled
posteriorly over the file; a pulse of sound is produced during
each slip Patek, 2001, 2002; Patek and Baio, 2007. The
plectrums are located at the base of each antenna and
traverse the oblong files located on each side of the antennu-
lar plate Fig. 1. Each rasp sound is produced when the
plectrum is pulled posteriorly and generates a series of sound
pulses as it sticks and slips over the surface of the file Fig.
2. Because there is a pair of plectrum/file units, sounds can
be produced by rubbing only one plectrum over the file, both
plectrums in series, or both plectrums concurrently. This
flexibility in the deployment of the sound-producing struc-
tures potentially offers spiny lobsters additional variation in
the range of signal features.
While the above studies have addressed the functional
morphology, evolutionary history, and behavioral context of
J. Acoust. Soc. Am., Vol. 125, No. 5, May 2009
sound production, the acoustics of these sounds have been
examined exclusively in laboratory settings and, thus, the
amplitude and frequency structure of these sounds in nature
are not currently known. Furthermore, little is known about
the influence of using paired structures on the signal features
and the scaling of signal features with body size in adult
lobsters Meyer-Rochow and Penrose, 1976; Patek and Oak-
ley, 2003. Examination of the sounds in the laboratory and
field offers important insights into the use of these signals by
spiny lobsters and, more generally, the role and function of
antipredator sounds in the marine environment. In this study,
as in all previous analyses of spiny lobster antipredator
sounds, we measure the rasps generated during handling,
simulating the lobsters experience once a predator has suc-
cessfully caught the prey and is attempting to process it; this
leaves open the possibility that a different suite of acoustic
signals are used during predator approach or for signaling to
distant predators, although no such sounds have been docu-
mented to date in spiny lobsters.
Thus, the goals of this study were to examine the acous-
tics and acoustic behavior of the California spiny lobster
Panulirus interruptus from the following three perspec-
tives: 1 Body size and signal features: Which acoustic pa-
rameters are correlated with body size? Do spiny lobsters
vary rasp duration by increasing number of pulses or de-
creasing pulse rate? 2 Plectrum activation and rasp varia-
tion: Are single, sequential and concurrent plectrum activa-
tion patterns correlated with specific rasp features, such as
greater rasp duration, higher pulse rate or greater number of
pulses, and particular behaviors, such as tail flipping or leg
movements? 3 Rasps and their acoustic environment: How
does the acoustic environment influence rasp signal features,
specifically when comparing recordings made in a tank ver-
sus in the field?
II. METHODS
A. Animal collection and care
California spiny lobsters, Panulirus interruptus Crusta-
cea, Decapoda, Palinuridae, were collected at the University
of Southern California, Wrigley Institute for Environmental
Studies WIES, Santa Catalina Island, CA, in baited lobster
FIG. 2. Color online Spiny lobsters typically produce
a series of rasps A each consisting of a series of
pulses B. In a rasp produced by a single plectrum the
first rasp in panel A, seven pulses are visible B. The
second rasp in A was generated with two plectrums ac-
tivated concurrently C, beginning with the first plec-
trum producing a series of pulses labeled a and the
second plectrum generating the overlapping series of
pulses labeled b. Series a and series b are distinguish-
able by differences in both amplitude and temporal
spacing.
Patek et al.: Spiny lobster acoustics 3435
traps or by hand CA Department of Fish and Game Permit
No. SC-5751. One small lobster was borrowed from the
WIES touch tank. Lobsters that were brought back to the
laboratory were maintained in cylindrical holding tanks
1.5 m diameter, 0.8 m height with a continuous supply of
sea water 14 16 C. They were fed bait fish daily.
Experiments were conducted during three different field
seasons, and the lobsters were captured shortly before each
set of experiments. In the first field season 2005, we con-
ducted the temporal acoustic analysis experiments in which
we recorded 24 individuals 2 males and 22 females;
44 102 mm carapace length; 14 15 C water temperature.
In the second field season 2006, we conducted audio-video
experiments of acoustic behavior and comparisons of rasp
acoustics in tank versus field conditions. In these experi-
ments, we recorded 20 individuals 6 males and 14 females;
50 113 mm carapace length; 8 18 mm file length;
14 15 C water temperature. In the third field season
2008, we recorded 13 more individuals 12 females, 1
male; 65 93 mm carapace length; 19.8 C water tempera-
ture in the field and measured pressure levels of the rasps
and ambient field environment. The specific sample sizes
used in each experiment are detailed below.
B. Temporal acoustic analyses
We tested for the presence of correlations between body
size and temporal components of the rasp, including rasp
duration s, number of pulses total number of pulses per
rasp, and average pulse rate pulses s1: number of pulses
per rasp divided by rasp duration Fig. 2. A hydrophone
1 Hz 170 kHz, TC4013, Reson, Slangerup, Denmark was
connected to a band-pass filter high-pass: 10 Hz, low-pass:
15 kHz; 1 Hz 1 MHz VP2000 voltage preamplifier, Reson,
Slangerup, Denmark and a digital audio recorder 48 kHz
sample rate, maximum 20 kHz frequency response
0.5 dB, PMD670, Marantz, NJ. Individuals were hand-
held at variable depths in a fiberglass tank 1.5 m diameter,
0.8 m height with the hydrophone suspended approximately
60 cm from the anterior end of the lobster.
Rasp waveforms were measured using acoustic software
RAVEN 1.2.1, Cornell Laboratory of Ornithology, NY. We
defined rasps as consisting of at least two pulses that oc-
curred within 45 ms of each other. Sometimes, a single pulse
of sound was produced in isolation; these pulses were not
included in the analyses. Pulse duration was not measured,
because previous studies have shown that tank reverberations
obscure the ending time of each pulse Patek and Baio,
2007. When the rasps were difficult to resolve against ex-
cessive background noise, they were omitted from the analy-
ses.
In many recordings, the sounds generated by the two
plectrums were distinguishable from each other, either tem-
porally or through amplitude differences. We split the dataset
into rasp waveforms unambiguously produced by one plec-
trum only and compared them to rasp waveforms clearly
showing activity from two plectrums Fig. 2; when this dis-
tinction was not clear, the rasps were not included in this
particular analysis. Using these distinct patterns, we docu-
3436 J. Acoust. Soc. Am., Vol. 125, No. 5, May 2009
mented three modes of plectrum activation: 1 single plec-
trum in which only one plectrum was used to generate one
rasp, 2 sequential plectrums in which the same plectrum
was used repeatedly or two plectrums were used in sequence
to generate multiple rasps, and 3 concurrent plectrums in
which two plectrums concurrently generated one or more
rasps see Fig. 2 and further explanation in Sec. III. We
limited this dataset to individuals that produced at least three
rasps each of single plectrum and dual plectrum activation.
We analyzed the relationships among these acoustic
variables and between these variables and body size. Least-
squares linear regressions were used to examine the correla-
tion between carapace length and mean values for acoustic
features across individuals. A general linear model analysis
of covariance ANCOVA was applied to examine the ef-
fects of plectrum activation on the temporal acoustic fea-
tures, as well as the effects of individual, and individual by
plectrum usage on the resulting correlations. Similarly, an
ANCOVA was used when examining the correlation between
pulse number and rasp duration within and across individu-
als. Statistics were performed with JMP v. 5.0.1 software.
C. Comparison of acoustic frequencies in tank
versus field conditions
The acoustic frequencies of P. interruptus rasps were
compared between field and tank recordings. Each individual
lobster was recorded in the field and then recorded in the
tank so that the spectral characteristics could be compared
both within and among individual lobsters.
Field recordings were taken in 7.3 m water 14 C with
the lobsters hand-held at 42 cm depth. The distance of the
hydrophone from the anterior end of the focal lobster was
held at a constant 31 cm in the field and ranged from
31 66 cm in the tank. The tank recordings were performed
in a cylindrical, fiberglass tank 1.5 m diameter, 0.8 m
height at 15 C. Calibrated recordings were taken with a
hydrophone 0.1 Hz 10 kHz 1.5 dB, sensitivity:
206.1 dB 0.25 dB re 1 V / Pa, Type 8104 hydrophone,
Brel and Kjaer, Nrum, Denmark and amplifier set at
high-pass filter 2 Hz and low-pass filter 10 kHz;
0.2 Hz 200 kHz, Type 2635 charge amplifier, Brel and
Kjaer, Nrum, Denmark which were connected to a digital
data acquisition system 50 kHz sample rate, NIDAQ 6062E
PCMCIA data acquisition card, National Instruments, TX;
custom data acquisition software, MATLAB, The Mathworks,
Natick, MA. Using a custom MATLAB program, the data
were converted to .wav files by scaling the voltage ampli-
tude by a factor of 0.1 and running a 20 Hz high-pass But-
terworth filter.
The dominant frequencies the two frequencies with
greatest acoustic power were identified for each rasp and
compared to ambient background noise in each recording
RAVEN v. 1.2.1 and 1.3, Cornell Laboratory of Ornithology,
Ithaca, NY. Temporal measurements were calculated from
the acoustic waveforms; frequency analyses were measured
from power spectra using a discrete Fourier transform set-
tings: Hanning window, 2000 sample window size, 3 dB fil-
ter bandwidth at 36 Hz resolution.
Patek et al.: Spiny lobster acoustics
 We examined correlations between acoustic features and
body size using least-squares linear regressions. We per-
formed a t-test to examine whether recording conditions sig-
nificantly affected the dominant frequencies. Statistics were
performed using JMP software v. 7.0, SAS Institute, Inc.,
NC.
D. Rasps in the field environment
The pressure levels of the rasps and background noise
were measured in the field. Lobsters were held by hand in
7.3 m water at 45 cm depth with the hydrophone positioned
at 97 cm depth. Thus, the effective diagonal distances be-
tween the lobster and hydrophone were 0.9, 1.1, 1.3, and
1.5 m. The equipment and settings were the same as in Sec.
II C. Absolute average power dB was calculated by con-
verting RAVEN softwares dimensionless units to pascals us-
ing the calibration provided by the hydrophone and amplifier
manufacturer and the conversion factors provided by RAVEN
software version 1.4, Hanning window, 2000 sample win-
dow size, 3 dB filter bandwidth at 36 Hz resolution. These
calibration methods are explained in the RAVEN software sup-
port documents and are also available upon request from the
authors. The average power dB was calculated relative to
1 Pa the standard for aquatic measurements and also cal-
culated relative to the baseline noise level measured in each
recording.
E. Audio-video analyses of acoustic behavior
In order to test whether rasp features and plectrum acti-
vation were correlated with specific behaviors, we used syn-
chronous audio and video to record spiny lobsters producing
rasps. Each individual was held approximately 36 cm deep,
1.5 m from the camera, and recorded until it produced 510
rasps 20 15 000 Hz; HTI-94-SSQ hydrophone, High Tech,
Inc., Gulfport, MS; Sony DCR-VX2100 Handycam video
camera, Tokyo, Japan; Amphibico VLAL0010 underwater
housing, Montreal, Canada. Rasps were elicited by holding
and gently squeezing or tickling the lobster.
We identified behavioral units typically associated with
escape or arousal in lobsters Atema and Cobb, 1980. The
two most consistent and identifiable behaviors were leg
movements and tail flips. Tail flips are an escape response in
which the tail is rapidly tucked under the body causing the
animal to rapidly jet backwards. Leg movements were noted
if they were vigorous and continuous as distinguished from
the slow or small movements associated with resting behav-
ior.
Sound production and behavioral units were counted and
binned over 10 s intervals. The onset of each 10 s bin oc-
curred when the spiny lobster first started to produce rasps.
Sound production and behavior were measured for 10 s in all
individuals with the exception of one individual for which
only 6 s were recorded. We tested whether including this
individual affected the results by running the analyses with
and without it. We logged the time at which each behavioral
unit and sound occurred and noted the identity of the plec-
trums right plectrum, left plectrum, or both producing the
sound. We then calculated the rate of rasp production the
J. Acoust. Soc. Am., Vol. 125, No. 5, May 2009
TABLE I. Temporal features of rasps. Sample size was 19 individuals with
521 rasps recorded per individual. The minimum number of pulses in a
rasp sequence was set at two pulses; single pulses were not included in the
analysis. A one-way analysis of variance tested for differences across indi-
viduals. ** indicates p 0.0001.
Minimum
maximum Mean std. dev. F-ratio
Pulse rate **7.9982
24192 71 20
pulses s1
Rasp duration
15303 108 35 **6.4656
ms
Number of
219 73 **8.5686
pulses
number of rasps divided by the 10 s bin during which they
occurred and the proportion of rasps produced by a single
plectrum or both plectrums including both sequential and
concurrent movement out of the total number of rasps pro-
duced during the 10 second time period.
We tested whether the behavioral units tail flip and an-
tennal movement were correlated with the rate of rasp pro-
duction and the number of rasps produced using both plec-
trums concurrently. These data were not normally distributed
Shapiro-Wilks Goodness-of-Fit test; p 0.0001, therefore
the nonparametric KruskalWallis test was used in place of a
t-test JMP 5.0.1, SAS Institute, Inc., NC.
Results are presented as mean one standard deviation.
III. RESULTS
A. Acoustic features and body size correlations
The temporal rasp features varied substantially both
within and across individuals Table I. File length Fig. 1
was positively correlated with carapace length Fig. 3 N
= 18; R2 = 0.6178, F = 25.8664, p = 0.0001. Carapace length
was negatively correlated with mean pulse rate, but not cor-
related with mean rasp duration or mean number of pulses
per rasp Fig. 3; Table II. Rasp duration was positively cor-
related with number of pulses Fig. 4 df = 18; whole model:
R2 = 0.6998, F = 24.7785, p 0.0001; number of pulses: F
= 181.4677, p 0.0001; individuals: F = 12.5963,
p 0.0001. Because pulse rate was calculated using values
from rasp duration and pulse number, it was not statistically
valid to examine the relationships among the three variables.
The rasps produced with two plectrums concurrently had a
significantly greater number of pulses than rasps produced
by one plectrum alone or two plectrums sequentially; rasp
duration and pulse rate were also greater in rasps produced
with two plectrums Table II.
Dominant frequencies were not correlated with body
size in the tank nor in the field Fig. 3 least-squares linear
regression, p 0.4 in all tests. Body size was not correlated
with average power when pooled across all recording dis-
tances Fig. 3 R2 = 0.302, df = 12, F = 4.76, p = 0.05.
B. Acoustic frequencies in the tank and the field
The frequency characteristics and background noise lev-
els of the rasp recordings were different in the field and the
Patek et al.: Spiny lobster acoustics 3437
 FIG. 3. The relationships between file
length A, temporal signal features
BD, spectrographic features
EF, and body size. Each data
point represents the mean value for an
individual lobster.

tank Table III. The rasps from the field recordings typically
had one distinct narrow peak below 500 Hz and another
broader peak around 1.5 2 kHz Fig. 5. Tank recordings
lacked this predictable structure and exhibited a pattern of
evenly spaced narrow peaks Fig. 5. The dominant frequen-
cies in the tank and field were significantly different
t-ratio= 7.569, p 0.0001, but the second most powerful
frequency was not significantly different t-ratio= 1.212, p
= 0.24.
C. Rasps in the field acoustic environment
The average power of the rasps was 150.4 2.0 dB re
1 Pa N = 13 lobsters; 281 rasps. The average background
noise level was 149.3 3.3 dB re 1 Pa N = 36 recordings
and the rasps exceeded the background noise level by an
average 1.6 dB range: 7.5 to 9.5 dB Table IV; Figs. 6
and 7. 31% of the rasps had power less than the average
ambient noise with the majority of the negative decibel ref-

TABLE II. The correlation between temporal rasp features, body size, and plectrum activation. Least-squares
linear regressions were used to analyze the relationship between pulse rate, rasp duration, number of pulses, and
carapace length. A general linear model ANCOVA was used to analyze the correlation between the use of one
or both plectrums to generate sound and the temporal features of the rasp. This second analysis was restricted
to individuals producing at least three rasps each of single plectrum and double plectrum activation, resulting in
a dataset of five individuals. * indicates p 0.05; ** indicates p 0.001.

Carapace
length
N = 19 Plectrum activation
individuals df = 1 , 4

Pulse rate *R2 = 0.2170 Whole model **R2 = 0.6501, F = 10.9405

F = 4.7110 plectrum activation *F = 4.1794
Individual **F = 18.6654
plectrum activation individual F = 0.2576

Rasp R2 = 0.0517 Whole model **R2 = 0.5289, F = 6.6105

duration F = 1.0308 plectrum activation **F = 13.5364
Individual **F = 9.4418
plectrum activation individual
F = 1.1452

Number R2 = 0.03826 Whole model **R2 = 5826, F = 8.2215

of pulses F = 0.6763 plectrum activation **F = 45.2203
Individual *F = 4.4095
plectrum activation individual F = 0. 1.9512

3438 J. Acoust. Soc. Am., Vol. 125, No. 5, May 2009 Patek et al.: Spiny lobster acoustics

FIG. 4. The number of pulses scales positively and significantly with rasp
duration. Each data point represents the mean value for an individual.
erenced to background noise rasps occurring at greater re-
cording distances Fig. 7. However, when the proportion of
rasps below zero dB re background noise was calculated
within each individual and then pooled across individuals for
each distance, this pattern was less evident and was non-
significant least-squares linear regression: R2 = 0.5532, df
= 1 , 12, F = 1.905, p = 0.09.
D. Audio-video analyses of acoustic behavior
Two datasets were analyzed, one including all data and
one excluding a short video sequence. The results were con-
sistent whether or not the short video clip was included in the
dataset; thus, the statistical results presented here include all
available data. The number of rasps produced by both plec-
trums concurrently Fig. 2c was positively correlated with
tail-flip behavior Table V. Regardless of whether rasps
were produced with a single plectrum or both, the rate of
sound production increased significantly when an individu-
als legs were moving Table V.
J. Acoust. Soc. Am., Vol. 125, No. 5, May 2009
TABLE III. Comparison of frequencies in the field versus tank recordings
and background noise versus rasps. Data are in the following format: mean
frequency s.d.; tank: N = 13 individuals 322 rasps per individual; field:
N = 11 individuals 534 rasps per individual.
Tank Field
Background Background
noise Rasp noise Rasp
Dominant
126 322 1794 338 366 706 633 374
frequency Hz
Second dominant
n.a. 1796 303 n.a. 1590 483
frequency Hz
IV. DISCUSSION
The acoustics of the California spiny lobsters rasp were
tied to the ambient environment, the individual behavior of
the lobsters, and, to a lesser extent, the size of the lobsters.
As we discuss below, the interconnections between the rasp
characteristics and the environment may be central to the
rasps function as an antipredator signal.
A. Body size and signal features
Although the size of the sound-producing apparatus was
tightly correlated with the body size of these animals, the
acoustic features were less so Fig. 3. Pulse rate was corre-
lated with body size, such that larger animals produced rasps
with a slower pulse rate Fig. 3d. However, dominant fre-
quency and power were not strongly associated with body
size Figs. 3e and 3f. Given that the rasps are broadband
signals with little tonal definition, it is perhaps unsurprising
that a significant correlation between body size and dominant
frequency was not observed. Future studies should examine a
broader range of body sizes and examine the effect of moti-
vation on signal features. For example, the stick-slip mecha-
nism of sound production may permit greater power output
when individuals pull the plectrum more tightly against the
file thereby generating a higher normal force and louder
FIG. 5. Recordings of the same lobster producing a
rasp in two different environments. In the field, the
pulse structure of the rasp is evident A and the sound
shows little resonant structure B power spectrum set-
tings as described in Sec. II. In a tank, the reverbera-
tions obscure the pulse structure C and a series of
harmonics are apparent D. The grayed spectra in B
and D indicate the signature of the ambient back-
ground noise. This particular lobster was positioned
53 cm from the hydrophone in the tank and 30 cm from
the hydrophone in the field, which may also have
caused spectral differences Akamatsu et al., 2002.
Patek et al.: Spiny lobster acoustics 3439
TABLE IV. Average power of rasps at varying distances in the field. Power
is reported as dB re 1 Pa and dB re background noise standard deviation.
Samples sizes are number of individuals N followed by number of rasps
recorded per individual.
0.9 m 1.1 m 1.3 m 1.5 m
N = 8 318 N = 11 322 N = 9 122 N = 5 19
dB re 1 Pa 150.2 1.4 150.2 2.1 149.2 1.6 149.2 1.7
dB re background 1.8 1.6 1.3 2.2 0.45 2.1 1.0 2.4
noise
sound Patek and Baio, 2007. The fact that the lobsters were
hand-held in this study may have elicited different signaling
behavior than in freely-moving individuals although their
rapid escape responses preclude measuring calibrated power
levels at known distances in freely-moving individuals.
Also, repeated stimulation of the same individuals may have
yielded habituation, again influencing signal feature patterns
over the time-course of these experiments. This might ex-
plain the unexpected variation in power levels across the four
field recording distances Table IV.
Scaling of sound with body size has been examined pre-
viously in several spiny lobster species and across the family
as a whole Meyer-Rochow and Penrose, 1974; 1976; Patek,
2002; Patek and Oakley, 2003; Patek and Baio, 2007.
Across the palinurid family, and within Panulirus argus,
pulse rate and number of pulses were positively correlated
FIG. 6. A rasp produced by a lobster in the field at 1.1 m from the hydro-
phone. The rasp is highlighted and is shown as a waveform upper and
spectrogram lower; RAVEN PRO software v. 1.4, Hanning window, 512
sample window size, 3 dB filter bandwidth at 140 Hz resolution. The
bracket indicates the energy extending below 1 kHz from the rasp, whereas
the ambient background noise is less powerful in this frequency range.
3440 J. Acoust. Soc. Am., Vol. 125, No. 5, May 2009
FIG. 7. The proportion of rasps with power levels above the background
noise black bars and below background noise level white bars at four
distances from the hydrophone. The width of the horizontal bars represents
the relative number of rasps recorded 0.9 m: 65 rasps; 1.1 m: 118 rasps;
1.3 m: 65 rasps; 1.5 m: 29 rasps. The overall mean is indicated to the right.
with the length of the file while rasp duration was negatively
correlated with file length Patek and Oakley, 2003. At early
developmental stages, correlations between dominant fre-
quencies and size were not found Meyer-Rochow and Pen-
rose, 1974; however, across juvenile and adult Panulirus
longipes, a positive correlation between body size and rasp
duration was found, and, similar to the results of this study,
there was a negative correlation between size and pulse rate
Meyer-Rochow and Penrose, 1976.
B. Plectrum activation and rasp variation
While the lack of body size and signal feature correla-
tion in spiny lobsters might be explained by the studys lim-
ited body size range, another key factor is the behavioral use
of the dual sound-producing apparatuses. For example, when
examining temporal features of rasps, a larger number of
pulses yielded a longer duration rasp Fig. 4. While this was
not explained by body size variation, it may instead be at-
tributed to the lobsters use the pair of plectrums rather than
a single plectrum to generate sounds.
Behavior, particularly the use of one or both plectrums
had strong influences on the temporal features of the rasp.
When two plectrums were used concurrently, the number of
pulses, rasp duration, and pulse rate were greater although
these results may be confounded by individual differences;
Table II. Furthermore, concurrent activation of both plec-
trums was correlated with the attempt to escape by tail-
flipping and the overall activity of the animal. Thus, the be-
TABLE V. Correlation between plectrum activation and behavioral units
tail flip and leg movement. A nonparametric, two-sample KruskalWallis
test was used to test whether these behaviors were associated with rasp rate
and plectrum use. Sample size was 20 individuals each sampled once. *
p 0.05.
Number of rasps when both
Overall rate of rasp plectrums were concurrently
production active
Tail flip Z = 0.6622; p = 0.51 Z = 2.3437; p = 0.019*
Leg movement Z = 2.2955; p = 0.022* Z = 1.7812; p = 0.075
Patek et al.: Spiny lobster acoustics
havioral motivation of the animal may more directly
influence signal characteristics than the body size even
though body size corresponds closely with the size of the
sound-producing apparatus.
The relevance of behavior to signal features has been
suggested previously in spiny lobsters Patek and Oakley,
2003 and demonstrated in other systems with multiple
signal-generating devices. For example, the searobin Pri-
onotus carolinus has a pair of sonic muscles which it can
contract simultaneously to generate greater amplitude or se-
quentially to produce a higher fundamental frequency Con-
naughton, 2004. The California mantis shrimp Hemisquilla
californiensis may also use its paired sonic muscles to vary
signal features Patek and Caldwell, 2006.
The behavioral manipulation of the signal features may
be important for tailoring an acoustic response to particular
predators. For example, multiple studies have shown that
vertebrates produce signal features specific to the predator
e.g., Templeton et al., 2005. Thus, it will be important in
future studies to present a range of predators to spiny lobsters
and assess whether they respond differently depending on the
relative size, risk, and hearing capabilities of that particular
predator.
C. Rasps and their acoustic environment
Consistent with previous studies Parvulescu, 1967;
Meyer-Rochow and Penrose, 1976; Akamatsu et al., 2002,
there were significant effects of the tank and field on the
frequency characteristics of the sound Fig. 5; Table III. The
tank recordings yielded average dominant frequencies of
1794 Hz, whereas in the field, the dominant frequencies av-
eraged 633 Hz. The second most powerful frequency was
similar in both settings1796 Hz in the tank and 1590 Hz in
the fieldsuggesting that the tank resonated the higher fre-
quencies in the rasp or damped the lower dominant fre-
quency. These substantial differences in frequencies and tem-
poral structure between the tank and field strongly suggest
that tank-based aquatic recordings should be interpreted with
caution and are not useful for comparisons and characteriza-
tions of frequency-spectra.
The high intensity collapse of cavitation bubbles domi-
nated the acoustic landscape around Santa Catalina Island,
the site of this study. The majority of these sounds in other,
similar environments have been attributed to snapping
shrimp Johnson et al., 1947; Au and Banks, 1998; Versluis
et al., 2000, although it is likely that cavitation sounds are
being produced by other organisms as well Colson et al.,
1998; Patek et al., 2004; Patek and Caldwell, 2005; Simon
et al., 2005. Consistent with our measurements of field
background noise averaging 149.3 dB re 1 Pa, snapping
shrimp Synalpheus paraneomeris generate signals at
183 189 dB re 1 Pa at 1 m from a hydrophone in a tank
Au and Banks, 1998.
The average power level of the rasps, 150 dB re 1 Pa,
is quite loud compared to measurements of marine acoustic
signals from similar sized organisms excluding the sound of
cavitation. A study of two spiny lobster species, Panulirus
homarus and Palinustus waguensis misspelled in the origi-
J. Acoust. Soc. Am., Vol. 125, No. 5, May 2009
nal paper, documented power levels of 50 143 dB Latha
et al., 2005; however, the reference level and distance from
the recording device were not specified, so it is difficult to
draw comparisons with the present data. The damselfish
Abudefduf abdominalis generates courtship calls at
105 119 dB re 1 Pa at 0.5 1 m Maruska et al., 2007.
Toadfish Halobatrachus didactylus acoustic power scales
with body size, with pressure levels ranging from approxi-
mately 108 140 dB re 1 Pa Vasconcelos and Ladich,
2008. Opsanus tau toadfish produce boatwhistle calls of an
average 126 dB re 1 Pa at 1 m Barimo and Fine, 1998.
Thus, the spiny lobsters rasp is loud, but so is the back-
ground noise Figs. 6 and 7. A primary consequence of the
loud background noise is that the rasps are obscured by the
ambient background noise even though they attenuate mini-
mally over the distances in which a predator encounter might
occur. Given that the rasps are similar in power to the ambi-
ent background noise, the probability that they will be ob-
scured is quite highapproximately 31% of the rasps re-
corded had a negative decibel level relative to the
background noise Fig. 7. This confers an advantage in the
context of the antipredator functionthe sounds are both
loud and local, and perhaps less likely to attract additional
nearby predators to the scene.
The frequency structure of the background noise relative
to the rasp may also be important for propagation Fig. 6. A
quiet window is present below 1 kHz, a region in which the
rasps power output is relatively high. It is possible that spiny
lobsters make use of such a window similar to gobies
shown to communicate in the quiet low-frequency region in
a noisy stream environment Lugli et al., 2003; Lugli and
Fine, 2007. Like the gobies, it is also possible that anti-
predator communication is occurring in the near-field, thus
measurements of particle velocity at close-ranges would
yield a more accurate portrait of the rasps acoustic land-
scape. While many of the spiny lobsters fish predators can
detect pressure waves, most marine organisms are also sen-
sitive to particle vibrations in the near-field. Characterizing
the near-field of these local antipredator sounds is necessary
both to understand the propagation of these signals and to
determine the relevant signal features to attacking predators.
Various lobster species have been shown to detect vibra-
tions in the near-field and at low frequencies less than
200 Hz, yet the presence of pressure-sensitive hearing struc-
tures in crustaceans remains contentious Cohen, 1955; Of-
futt, 1970; Tazaki and Ohnishi, 1974; Goodall et al., 1990;
Budelmann, 1992; Popper et al., 2001; Lovell et al., 2005;
Lovell et al., 2006. Previous research suggested that the
rasps could function in the near-field to warn neighboring
conspecifics Lindberg, 1955; Meyer-Rochow et al., 1982
via an alarm signal. However, given that palinurid larvae
cycle for many months before settling Phillips et al., 2006,
it is unlikely that they are genetically related and thus the
fundamental assumption that alarm calls aid close relatives
Caro, 2005 would not be met.
In conclusion, there is a web of interconnections be-
tween the basic mechanism of sound production, the behav-
ioral deployment, and the ambient environment in which
these sounds are produced; each component is essential to
Patek et al.: Spiny lobster acoustics 3441
the production and propagation of the signal. Across the size
ranges of the lobsters included in this study, there are no
obvious signals to potential predators about body size. How-
ever, the overall behavior of the lobster strongly impacts the
rasp features produced, suggesting that there may be a more
important association between the stimulus and acoustic re-
sponse than we addressed in these particular experiments.
The loud natural environment plays a key role in masking
the rasps even though the aquatic environment minimally
attenuates sound over these small distances. This first analy-
sis of antipredator acoustics and behavior in the California
spiny lobster suggests that much remains to be learned in this
rich frontier of marine bioacoustic research.
ACKNOWLEDGMENTS
The authors thank J. Baio, M. deVries, R. Jaciento, B.
Nowroozi, M. Patek, V. Patek, A. Summers, N. Valencia, and
W. Van Trump for helping to collect the sound and video
recordings. They thank two anonymous reviewers for their
constructive comments. They appreciate the staff support by
the Wrigley Institute for Environmental Studies. Funding to
SNP was provided by the UC Berkeley Undergraduate Re-
search Apprenticeship Program, the Committee on Research
Junior Faculty Research Grant and the Hellman Family Fac-
ulty Fund.
Akamatsu, T., Okumura, T., Novarini, N., and Yan, H. 2002. Empirical
refinements applicable to the recording of fish sounds in small tanks, J.
Acoust. Soc. Am. 112, 30733082.
Alexander, A. J. 1958. On the stridulation of scorpions, Behaviour 12,
333352.
Atema, J., and Cobb, J. S. 1980. Social behavior, in The Biology and
Management of Lobsters: Physiology and Behavior, edited by J. S. Cobb
and B. F. Phillips Academic, New York, pp. 409450.
Athenaeus 300. Le Banquet des Savants [The Banquet of Scientists] Lamy
Ed., Paris.
Au, W. W. L., and Banks, K. 1998. The acoustics of the snapping shrimp
Synalpheus parneomeris in Kaneohe Bay, J. Acoust. Soc. Am. 103, 41
47.
Barimo, J. F., and Fine, M. L. 1998. Relationship of swim-bladder shape
to the directionality pattern of underwater sound in the oyster toadfish,
Can. J. Zool. 76, 134143.
Bouwma, P. E. 2006. Aspects of antipredation in Panulirus argus and
Panulirus guttatus: Behavior, morphology and ontogeny, in Biological
Science Florida State University, Tallahassee.
Bouwma, P., and Herrnkind, W. 2004. The antipredator function of sound
production by the spiny lobster Panulirus argus, in Seventh International
Conference and Workshop on Lobster Biology and Management Hobart,
Tasmania.
Bouwma, P., and Hernnkind, W. 2007. Aspects of antipredation in Panu-
lirus argus and P. guttatus: Behavior, morphology and ontogeny, The
Lobster Newsletter 20, 46.
Budelmann, B. U. 1992. Hearing in Crustacea, in The Evolutionary Bi-
ology of Hearing, edited by D. B. Webster, R. R. Fay, and A. N. Popper
Springer-Verlag, New York, pp. 131139.
Caro, T. 2005. Antipredator Defenses in Birds and Mammals The Univer-
sity of Chicago Press, Chicago.
Chivers, D. P., Brown, G. E., and Smith, R. J. F. 1996. The evolution of
chemical alarm signals: Attracting predators benefits alarm signal send-
ers, Am. Nat. 148, 649659.
Cohen, M. J. 1955. The function of receptors in the statocyst of the
lobster Homarus americanus, J. Physiol. London 140, 934.
Colson, D. J., Patek, S. N., Brainerd, E. L., and Lewis, S. M. 1998. Sound
production during feeding in Hippocampus seahorses Syngnathidae,
Environ. Biol. Fishes 51, 221229.
Connaughton, M. A. 2004. Sound generation in the searobin Prionotus
carolinus, a fish with alternate sonic muscle contraction, J. Exp. Biol.
3442 J. Acoust. Soc. Am., Vol. 125, No. 5, May 2009
207, 16431654.
Edmunds, M. 1974. Defence in Animals: A Survey of Anti-Predator De-
fenses Longman Group Ltd., Essex.
Field, L. H., Evans, A., and MacMillan, D. L. 1987. Sound production
and stridulatory structures in hermit crabs of the genus Trizopagurus,
Journal of Marine Biology, U.K. 67, 89110.
Fish, M. P., and Mowbray, W. H. 1970. Sounds of Western North Atlantic
Fishes The Johns Hopkins Press, Baltimore.
Gamberale, G., and Tullberg, B. S. 1996. Evidence for a peak-shift in
predator generalization among aposematic prey, Proc. R. Soc. London,
Ser. B 263, 13291334.
George, R. W., and Main, A. R. 1967. The evolution of spiny lobsters
Palinuridae: A study of evolution in the marine environment, Evolution
Lawrence, Kans. 21, 803820.
Gittleman, J. L., and Harvey, P. H. 1980. Why are distasteful prey not
cryptic?, Nature London 286, 149150.
Goodall, C., Chapman, C., and Neil, D. 1990. The acoustic response
threhold of the Norway lobster, Nephrops norvegicus L. in a free sound
field, in Frontiers in Crustacean Neurobiology, edited by K. Wiese, W.
D. Krenz, J. Tautz, and H. Reichert Birkhauser, Boston, pp. 106113.
Guinot-Dumortier, D., and Dumortier, B. 1960. La stridulation chez les
Crabes Stridulation by crabs, Crustaceana 2, 117155.
Henninger, H. P., and Watson, W. H., III 2005. Mechanisms underlying
the production of carapace vibrations and associated waterborne sounds in
the American lobster, Homarus americanus, J. Exp. Biol. 208, 3421
3429.
Johnson, M. W., Everest, F. A., and Young, R. W. 1947. The role of
snapping shrimp Crangon and Synalpheus in the production of underwa-
ter noise in the sea, Biol. Bull. 93, 122138.
Latha, G., Senthilvadivu, S., Venkatesan, R., and Rajendran, V. 2005.
Sound of shallow and deep water lobsters: Measurements, analysis and
characterization L, J. Acoust. Soc. Am. 117, 27202723.
Lewis, E. E., and Cane, J. H. 1990. Stridulation as a primary antipredator
defense of a beetle, Anim. Behav. 40, 10031004.
Lindberg, R. G. 1955. Growth, population dynamics, and field behavior
in the spiny lobster, Panulirus interruptus Randall, University of Cali-
fornia Publications in Zoology 59, 157248.
Lovell, J. M., Findlay, M. M., Moate, R. M., and Yan, H. Y. 2005. The
hearing abilities of the prawn Palaemon serratus, Comp. Biochem.
Physiol. A 140, 89100.
Lovell, J. M., Moate, R. M., Christiansen, L., and Findlay, M. M. 2006.
The relationship between body size and evoked potentials from the sta-
tocysts of the prawn Palaemon serratus, J. Exp. Biol. 209, 24802485.
Lugli, M., and Fine, M. L. 2007. Stream ambient noise, spectrum and
propagation of sounds in the goby Padogobius martensii: Sound pres-
sure and particle velocity, J. Acoust. Soc. Am. 122, 28812892.
Lugli, M., Yan, H. Y., and Fine, M. L. 2003. Acoustic communication in
two freshwater gobies: The relationship between ambient noise, hearing
thresholds and sound spectrum, J. Comp. Physiol. 189, 309320.
Maruska, K. P., Boyle, K. S., Dewan, L. R., and Tricas, T. C. 2007. Sound
production and spectral hearing sensitivity in the Hawaiian sergeant dam-
selfish, Abudefduf abdominalis, J. Exp. Biol. 210, 39904004.
Masters, W. M. 1979. Insect disturbance stridulation: Its defensive role,
Behav. Ecol. Sociobiol. 5, 187200.
Mendelson, M. 1969. Electrical and mechanical characteristics of a very
fast lobster muscle, J. Cell Biol. 42, 548563.
Meyer-Rochow, V. B., and Penrose, J. D. 1974. Sound production and
sound emission apparatus in puerulus and postpuerulus of the western rock
lobster Panulirus longipes, J. Exp. Zool. 189, 283289.
Meyer-Rochow, V. B., and Penrose, J. D. 1976. Sound production by the
western rock lobster Panulirus longipes Milne Edwards, Journal of Ex-
perimental Marine Biology and Ecology 23, 191209.
Meyer-Rochow, V. B., Penrose, J. D., Oldfield, B. P., and Bailey, W. J.
1982. Phonoresponses in the rock lobster Panulirus longipes Milne
Edwards, Behav. Neural Biol. 34, 331336.
Moulton, J. 1957. Sound production in the spiny lobster Panulirus argus
Latreille, Biol. Bull. 113, 286295.
Mulligan, B. E., and Fischer, R. B. 1977. Sounds and behavior of the
spiny lobster Panulirus argus, Crustaceana 32, 185199.
Offutt, C. G. 1970. Acoustic stimulus perception by the American lobster
Homarus americanus Decapoda, Experientia 26, 12761279.
Parker, T. J. 1878. Note on the stridulation organ of Panulirus vulgaris,
Proceedings of the Zoological Society of London 1878, 442444.
Parker, T. J. 1883. On the structure of the head in Palinurus with special
Patek et al.: Spiny lobster acoustics
 reference to the classification of the genus, Nature London 29, 189
190.
Parvulescu, A. 1967. The acoustics of small tanks, in Marine Bioacous-
tics, edited by W. N. Tavolga Pergamon, New York, pp. 79.
Patek, S. N. 2001. Spiny lobsters stick and slip to make sound, Nature
London 411, 153154.
Patek, S. 2002. Squeaking with a sliding joint: Mechanics and motor
control of sound production in palinurid lobsters, J. Exp. Biol. 205,
23752385.
Patek, S. N., and Baio, J. E. 2007. The acoustic mechanics of stick-slip
friction in the California spiny lobster Panulirus interruptus, J. Exp.
Biol. 210, 35383546.
Patek, S. N., and Caldwell, R. L. 2005. Extreme impact and cavitation
forces of a biological hammer: Strike forces of the peacock mantis shrimp
Odontodactylus scyllarus, J. Exp. Biol. 208, 36553664.
Patek, S. N., and Caldwell, R. L. 2006. The stomatopod rumble: Sound
production in Hemisquilla californiensis, Mar. Freshwater Behav.
Physiol. 39, 99111.
Patek, S. N., and Oakley, T. H. 2003. Comparative tests of evolutionary
tradeoffs in a palinurid lobster acoustic system, Evolution Lawrence,
Kans. 57, 20822100.
Patek, S. N., Korff, W. L., and Caldwell, R. L. 2004. Deadly strike
mechanism of a mantis shrimp, Nature London 428, 819820.
Patek, S. N., Feldmann, R. M., Porter, M., and Tshudy, D. 2006. Phylog-
eny and evolution of lobsters, in Lobsters: Biology, Management, Aquac-
ulture and Fisheries, edited by B. F. Phillips Blackwell, Ames, IA.
Phillips, B. F., Booth, J. D., Cobb, J. S., Jeffs, A. G., and McWilliam, P.
2006. Larval and postlarval ecology, in Lobsters: Biology, Manage-
ment, Aquaculture and Fisheries, edited by B. F. Phillips Blackwell, Ox-
ford, pp. 231262.
Popper, A. N., Salmon, M., and Horch, K. W. 2001. Acoustic detection
and communication by decapod crustaceans, J. Comp. Physiol., A 187,
8389.
Ruxton, G. D., Sherratt, T. N., and Speed, M. P. 2004. Avoiding Attack:
J. Acoust. Soc. Am., Vol. 125, No. 5, May 2009
The Evolutionary Ecology of Crypsis, Warning Signals and Mimicry Ox-
ford University Press, Oxford.
Sandeman, D. C., and Wilkens, L. A. 1982. Sound production by abdomi-
nal stridulation in the Australian Murray River crayfish, Euastacus arma-
tus, J. Exp. Biol. 99, 469472.
Sargent, T. D. 1990. Startle as an anti-predator mechanism, with special
reference to the underwing moths, Catocala, in Insect Defenses: Adap-
tive Mechanisms and Strategies of Prey and Predators, edited by D. L.
Evans and J. O. Schmidt State University of New York Press, Albany,
pp. 229249.
Sherratt, T. N., and Beatty, C. D. 2003. The evolution of warning signals
as reliable indicators of prey defense, Am. Nat. 162, 377389.
Simon, M., Wahlberg, M., Ugarte, F., and Miller, L. A. 2005. Acoustic
characteristics of underwater tail slaps used by Norwegian and Icelandic
killer whales Orcinus orca to debilitate herring Clupea harengus, J.
Exp. Biol. 208, 24592466.
Smale, M. 1974. The warning squeak of the Natal rock lobster, South
African Association Marine Biology Research Bulletin 11, 1719.
Speed, M. P. 2000. Warning signals, receiver psychology and predator
memory, Anim. Behav. 60, 269278.
Summers, A. 2001. The lobsters violin, Natural History 110, 2627.
Tazaki, K., and Ohnishi, M. 1974. Responses from tactile receptors in the
antenna of the spiny lobster Panulirus japonicus, Comp. Biochem.
Physiol. A 47A, 13231327.
Templeton, C. N., Greene, E., and Davis, K. 2005. Allometry of alarm
calls: Black-capped chickadees encode information about predator size,
Science 308, 19341937.
Vasconcelos, R. O., and Ladich, F. 2008. Development of vocalization,
auditory sensitivity and acoustic communication in the Lusitanian toadfish
Halobatrachus didactylus, J. Exp. Biol. 211, 502509.
Versluis, M., Schmitz, B., von der Heydt, A., and Lohse, D. 2000. How
snapping shrimp snap: Through cavitating bubbles, Science 289, 2114
2117.
Patek et al.: Spiny lobster acoustics 3443